|
|
|
|
|
|
by cocomutator
1615 days ago
|
|
|
I wonder how a modern language model would fare here -- use something like GPT-3 to evaluate the log likelihood gain of stitching together each of all N^2 possible pairs, then merge greedily best matches until none are left. Totally within reach, I bet it could get at least _some_ of the order right. |
|
|